15 July 2026 · 14 min read

The Power of Organizational Cultural Health - In Numbers

Culture is the most consequential variable in how a group of people performs, and it is the one most often argued for with adjectives. Here it is with numbers instead - including the ones that cut against us.

Says who?

Leila has eleven minutes with the board on Thursday and one slide about culture. It has a good number on it, borrowed from a deck somebody forwarded her, and she has practised saying it without looking down.

The finance director reads the slide, takes off her glasses and asks the only question finance directors ever really ask. "Says who?"

Leila does not know. The deck cited a blog. The blog cited a report. When she goes looking that night, the report says something narrower, or is not where it was said to be, or was never written. The money goes to something with a receipt.

That is where most culture arguments die, and they deserve to. So this piece does what the finance director did. Every figure names its source, and one section below says how far you could lean on each of them in that room.

The upside of organisational culture, measured

Engagement. Global engagement fell to 20% in 2025 - the lowest since 2020, and a second consecutive annual decline (Gallup, State of the Global Workplace 2026). One person in five engaged by Gallup's measure leaves most of the available performance in a workforce on the table.

Satisfaction. People who rate their workplace culture good or excellent are 790% more likely to be satisfied at work than those who rate it poor or terrible (SHRM, 2023).

The cost of the gap. Gallup puts the cost of low engagement at about $10 trillion a year in lost productivity, roughly 9% of world GDP (State of the Global Workplace 2026). Whatever a culture programme costs, it is being weighed against that.

Performance. Teams with generative cultures had 30% higher organisational performance than those without, in research drawing on more than 36,000 professionals over nine years (DeBellis and Harvey, "2023 State of DevOps Report: Culture is everything", Google Cloud, 5 October 2023). The report means by generative that people felt included and like they belonged on their team. Across 136 independent samples covering more than 22,000 people, psychological safety is linked with how well people perform their tasks (Frazier, Fainshmidt, Klinger, Pezeshkan and Vracheva, Personnel Psychology, 2017).

The next bet. Culture, manager support and talent practices explain 67% of the impact AI actually has at work - more than twice individual mindset and behaviour (32%) (Microsoft, Work Trend Index 2026, 5 May 2026). A company that stops at teaching individuals to use the tools has, by Microsoft's own reckoning, paid for the smaller half.

Employee engagement: watch the managers' number

The 20% is the figure everybody prints. The one under it is more alarming. Only 22% of managers worldwide were engaged in 2025, down from 27% a year earlier and nine points below 2022 (Gallup, State of the Global Workplace 2026).

That matters because of where managers sit. Line management, leadership, skills development and meaningful work have the biggest effect on all five business-related outcomes the CIPD measures (CIPD, Good Work Index 2026, July 2026). So the first name on that list is a group whose engagement dropped five points in a year.

Picture a team lead at ten to nine on a Monday. She stopped believing in the plan some time in the spring, and at nine she has to present it to six people who will take their cue from her face. Every value on the wall reaches those six through her. Whatever the company spends on engagement, it passes through that Monday first.

The downside, measured

The flip side is the half most vendors skip, and it is the more useful half.

People made happier worked about 12% more productively than a control group (Oswald, Proto and Sgroi, University of Warwick, 2015). They were adding numbers for pay in a laboratory rather than doing their jobs, so read it as a direction rather than as a workplace estimate.

Run the same thought in reverse and it stops being about a laboratory. 964,000 workers in Great Britain had stress, depression or anxiety caused or made worse by work in 2024/25 (HSE, Key figures for Great Britain 2024 to 2025, updated 30 June 2026). 45% of UK employees feel unable to raise a mistake or a risk at work, and 15% say they have made a preventable mistake because they felt unsafe speaking up (Mental Health First Aid England, The cost of workplace silence, 2 February 2026). Of US workers who had met a toxic workplace, 79.1% blamed unethical, unaccountable or unsupportive leadership; 68.9% had worked for an employer they consider toxic, and 47.6% had quit a job because of one (iHire, 2026 Toxic Workplace Trends Report, PR Newswire, 25 June 2026).

Those three cannot be added together: different countries, different questions. They are worth setting side by side anyway, because each one is a person. Somebody signed off sick. Somebody saw the wrong figure on Tuesday, looked at their manager's door and shut the file. Somebody handed in their notice and told everyone it was the commute.

How to read culture statistics before a board does

A note on confidence, because it matters more than the figures: every number here names its source, and several that circulate widely in culture writing are missing on purpose. An engagement gap credited to Deloitte, a turnover split credited to Columbia University and an absenteeism figure credited to the APA could not be found at the source they are credited to, so they are not here. If you are putting a number in front of a board, check the primary source rather than trusting a blog - including this one.

The best-known number in change management is the cautionary tale. A critical review of five published instances of the claim that 70 per cent of organisational change initiatives fail found no valid and reliable empirical evidence for it (Mark Hughes, Journal of Change Management 11(4), 2011). You will still see it on a slide this week.

Here is how far we would lean on the figures above, in Leila's room.

A survey tells you what people said

Gallup's 20% and SHRM's 790% are people describing their own working lives. The 790% also compares the top of a rating scale with the bottom, with both answers coming from one person, so it is a fair sign that people who like their culture like their job, and no evidence that one produced the other. Say "people report", and you will survive the follow-up question.

A correlation tells you two things move together

The 136 samples and the 30% are links. A team that is doing well may feel safer because it is doing well. A great many studies pointing the same way make a link worth acting on. They do not turn it into a cause, and a good finance director will say so before you do.

A laboratory tells you which way something points

The 12% came from people adding up numbers. It is good evidence that mood and output are connected in that room, and it cannot be multiplied by your payroll.

A modelled cost tells you the size of the argument

The $10 trillion is Gallup's estimate of a whole world. It is a fair way to say the problem is large. It says nothing about what low engagement costs you, and anybody who divides it by your headcount is doing arithmetic rather than analysis.

A source with something to sell is still a source

Several of these figures come from companies in this market, and this page is written by another one. Keep the number, name who found it, and ask what they would have printed if the answer had gone the other way.

Does a good culture make more money?

Not always, and not straight away. Here is the number that cuts against us. A third of chief executives (33%) describe their own workplace as toxic, up 25 points in a year, and 40% of employees say the same, up 18 points; executives in toxic cultures were about twice as likely to report strong financial growth (70% against 36%) (Businessolver, 23 June 2026).

Most culture writing would leave that out, and we would rather you heard it from us. It is 300 US executives describing their own year, it is one year, and it is no proof that cruelty pays. Businessolver's own headline says toxic workplaces are winning "for now". But it does show that a bad culture can sit beside a good year, which kills the easy version of the argument, where culture is a lever you pull in January and cash in December.

The honest case is the one a surveyor makes about a roof. It does not leak in a dry year, and that tells you nothing about the roof.

The findings that argue against the obvious fix

This is where most culture writing stops, and where it gets interesting.

Wellbeing programmes mostly do not work. The largest study of individual workplace wellbeing interventions to date - 46,336 workers, 233 UK organisations (William Fleming, University of Oxford, 2024) - found participants in mindfulness apps, resilience classes and wellbeing programmes no better off than colleagues who skipped them.

Training decays. Across 42 samples in 37 studies of workplace resilience programmes (Vanhove and colleagues, 2016), the effect fell from d = 0.26 immediately after training to 0.07 at follow-up. Not "training does not work" - training works, and then stops.

Fixes come back. Just 6% of organisations said behaviour improved permanently following intervention, according to research by WorkNest. In a WorkNest survey of 1,017 HR professionals and business leaders, 45% experienced an initial improvement before the problematic behaviour returned (WorkNest, via HRreview, 21 August 2026). That is the same curve as Vanhove's, seen from a manager's desk.

The list on the wall predicts very little. Across 562 companies that published their values, there was no correlation between the values they emphasised and how their employees rated the culture (Sull, Turconi and Sull, MIT Sloan Management Review, 21 July 2020).

The surveys are exhausted. The instrument is worn out; the intent behind it is not.

Signal-only rituals die. A daily check-in that collects a mood and gives the employee nothing back is a small tax, charged every morning. The lesson is precise: a daily ritual has to give something first, and what it has to give back is an article of its own.

What the numbers actually recommend

Read together, the evidence is not "culture matters, so buy a culture product". It is narrower and more awkward than that:

• The effect is real and large, so it is worth spending on.
• One-shot interventions decay to nothing, so cadence beats intensity.
• Asking people to rate their culture is a spent instrument, so the measurement has to come from behaviour rather than from a form.
• A ritual that only extracts will not survive contact with a busy week.
• A good year proves nothing about the culture under it, so the case cannot rest on the last quarter.

How MeaningFuel helps with cultural health

Those numbers are the entire design brief, and they are why the thing we built is two minutes rather than two days: one insight, one quote behind it, four faces, one question, every working day, with an insight underneath it, written per theme. Nobody receiving it is handed a form: one short optional question at most, every 14, 30 or 90 days, or never as the employer sets it, pooled with nobody named. What gets read is whether the habit is alive: how engaged people are, how many take part, and whether it landed.

The daily insight

The company chooses the values and skills it wants to work on, and each working day everyone gets one short insight on one of them. Where these numbers point, that means Purpose, Trust and Resilience among the values, and Leadership & Coaching for the managers in that 22%. In an ordinary week a reader might meet a line on what an ordinary task is for, and another on what a manager owes the person who brings bad news. Low engagement itself cannot be chosen for the daily insight, because it is a problem rather than something a company holds. That is what a module is for.

A module for a team that has gone flat

Where a specific problem shows up, a module, two working weeks of short daily lessons on that problem, gets booked onto a team - never onto a person, because culture is a property of a group and because the alternative is surveillance with better branding.

Low engagement tends to arrive as a team that has gone flat, and the module for that is A Motivated Team. Booked for a team, it invites the people on it, and those who join get one short lesson each working day. It treats low morale as disconnection rather than laziness: from progress, from the people the work reaches and from choice. So the team marks small wins where they can be seen, traces a task to the person it serves, and each person picks a stretch of their own. Where the real cause is concrete, such as pay or a broken promise, the module says so and the team takes it upward, because no amount of practice covers for that.

What leaders read

Insights shows how much of a team takes part and how the insights land, pooled, with nobody named. That would tell Leila whether the habit is alive in the team she is worried about, and whether the insights are landing or merely arriving. It would not give her a return on investment, and we would be wary of anything that claimed to.

It does not promise a better engagement score, and it will not bring back a manager who stopped believing in the spring. It keeps the company's values in front of people every working day, which is the one thing the numbers above say a one-off cannot do.

Questions a board will ask about culture

What is cultural health?

Cultural health is how closely what a company says it values matches what people actually do on an ordinary day. It shows in behaviour: whether bad news travels, whether people take part, whether a value survives a hard quarter. One good survey score is not the same thing, because the score records how people felt about the question on the day they answered it.

How do you measure organisational culture?

Measure what people do rather than what they say they feel. Participation over time, how widely it spreads across teams and whether people say the material landed all tell you more than a once-a-year rating, because behaviour is harder to give a polite answer to. A survey still has a place, as one reading rather than as the instrument.

What is the return on a culture programme?

Nobody can tell you honestly, so distrust anybody who offers a single figure. The evidence supports a direction: better cultures go with better engagement, performance and satisfaction. It does not support a multiple you can put in a budget. The defensible case is what the alternatives cost, and how little a one-off leaves behind.

Is employee engagement the same as culture?

No. Engagement is one reading of how people feel about their work, and culture is the set of habits that produces the feeling. Engagement can dip in a bad month while the culture holds, and a good year can flatter both. Track it, but treat it as a symptom rather than the thing itself.

The number to take to the board

Which of these numbers is the one worth acting on depends on what you are trying to defend. If it is budget, it is the $10 trillion. If it is the approach, it is Vanhove and colleagues' d = 0.26 decaying to 0.07 - because that one says the problem was never the method.

Either way, have the answer ready. "Says who?" takes four seconds to ask.

• Organisational Culture
• Employee Engagement
• Workplace Culture
• Research
• Retention